1\section{Conclusion}2\label{sec:conclusion}34This paper has built a reproducible pipeline from \nRecordsIdentifiedRaw{} bibliographic records to \nUsableEstimates{}5harmonised implicit prices from \nStudiesExtracted{} hedonic studies in \nCountriesExtracted{} countries, and has used it to answer6the questions the Sirmans programme posed in 2005 with the tools of 2026. The robust findings are few and clear. The structural7core of the hedonic model is global: living area, bathrooms, age and distance to the centre have implicit prices whose means are8precisely estimated and similar across continents and decades once the specification is controlled. The signature amenities9carry premiums of the order documented in the attribute-specific literature. Every implicit price is nonetheless heterogeneous10beyond sampling error, with prediction intervals that span zero for most characteristics, and the observable features of a11study, including the estimator, explain little of that heterogeneity. The tentative findings are the geographic and temporal12patterns, which are confounded by the composition of the literature; the publication-selection diagnostics, which are13suggestive for amenity premiums but not interpretable as corrections; and the AI evidence, which indicates better prediction14from tree ensembles and modest incremental information from text and image features in a small sample. The exploratory findings15are the machine-learning meta-regressions, which do not predict implicit prices out of study. The paper's contribution relative16to \citet{sirmans2005composition,sirmans2006value} is therefore a global, harmonised and uncertainty-aware restatement of what the17hedonic literature knows about the value of housing characteristics, together with a documented account of what it cannot know18until the literature reports its results in a form that can be accumulated.19